Are all start capacitors equal ?

[JacquesBBB]

My friend has an old electric   grinder with a AC monophase  motor
which could not start unless you give it  a little push.

I checked the capacitor (4uF, 220V)  which dated from  1968, and
which of course is completely dead.

I had no such capacitors but put  4  4.7uF , 160V  polypropylene caps  as 2x2  in order to get 4.7uF and 320V.
I had check theses caps which had only 0.2 Ohm ESR and nominal value for the capacity.

I had some good results like that on another  old motor, but this time it did not do anything,
and the engine still does not start alone.

I check the resistor of the engine and verified that  there was no cuts in any of the two
wirings.

So here is my questions :

- Are all capacitors equal for use as start capacitors ? (of course with no polarity).

- Is the capacitor value important ( can I put a larger value without any problem) ?

- What would you do to start this motor ?

Thanks

[Ian.M]

The capacitor is matched to the starting winding for the best combination of current and phase shift to get the maximum starting torque.   Too big a cap and you get a higher current but low phase shift, too small a cap and you get lots of phase shift but little current.

If its got a centrifugal switch, the chances are that's gone bad and the starting winding isn't cutting in.

See http://www.repairfaq.org/REPAIR/F_appfaqh.html#APPFAQH_003

[sdg]

- What would you do to start this motor ?

I'd check just about everything: brushes, switch, bearings & angle gears, etc. Dismantle, clean, lube, replace what needs to be replaced...
Or buy a new one 

[JacquesBBB]

@ Ian.M

thanks for the link. It looks very interesting. I could not find such synthetic information on the web.

@sdg :

"buy a new one" is not an option. This is a vintage motor  that I want to keep as much as possible in its original state.

[SeanB]

As mentioned above you probably have a centrifugal start switch ( and the associated capacitor) that is open circuit on the switch, or which has worn contacts, or a worn coupling link for the switch.

[Seekonk]

Did you check the replacement capacitor(s) after the test? Capacitors are not equal. The welds are the weakest part and can open with high current.  Film capacitors are common in that range (ceiling fans) but 4 uF seems like a low value for a grinder.

[KJDS]

If you put two caps in series then unless you add in bleed resistors it's likely that most of the voltage gets dumped across one of them.

[JacquesBBB]

If you put two caps in series then unless you add in bleed resistors it's likely that most of the voltage gets dumped across one of them.

Please provide a reference for this.

[SeanB]

Only an issue if they are run at close to max voltage and then the typical spread in production tolerance of 10% means the lowest value device has the highest voltage across it. With those film caps they likely will not be an issue, they will just go bang.

My bets are still on the start switch being open, or there is a 100uF start capacitor there as well that has died.

Photos will help.

[Zero999]

If you put two caps in series then unless you add in bleed resistors it's likely that most of the voltage gets dumped across one of them.

Please provide a reference for this.

None is needed. Work it out.

If each capacitor has a tolerance of say 20% then it can be calculated using the potential divider formula.

This is more of an issue at DC though as the leakage current dominates and is less predictable.

[Totalsolutions]

"What would you do to start this motor" - fit the correct rated cap. Its done well for the age and has lost value over the years. Stand back when it starts.

[Zero999]

What was the failure mode for the old capacitor? Did it read short circuit? If so it's likely it's overheated some of the motor windings.

[JacquesBBB]

None is needed. Work it out.

OK, if I understand well, the problem could be that the two leakage resistors are very different.

( I must say that I have tested all capacitors values and they were very similar).

The bleeding resistors should thus be equal, with values smaller than the lowest leakage resistor.
I will measure this  tomorrow evening when I will go back to my workshop. 
To be a problem, I suppose that one leakage resistor needs to be less than half of the other one, so one would
have < 80 V and the other one > 160 V.
I assume that 120 kohm resistors should be OK ? (after verification of the leakage resistors ).


I will also verify the state of the  caps after the test, which I did not do.

I will have also to see if there is a centrifugal start switch. I have not yet dismounted everything,
as  I expected I could just  start it by changing the cap.

[JacquesBBB]

As the full computation  ( two capacitors in series with two bleeding resistors, forced by a sinusoidal voltage),
is quite involve, even if one looks only to the stationary state,

I just made the  simulation on Everycircuit, and realize that unless you use very unrealistic small values for the
leakage resistors, the voltages at the caps are merely given by the computation without
leakage resistors ,
that is  V1*C1 = V2*C2,
So in the present case, as the two caps are practically identical,  the two  voltages are the same, and nearly equal
to 120 V .

But the peak voltage should be  120 * 1.4 = 168 V   so little bit beyond the mentioned 160 V.



But  when a capacitor is given for 160 V,  what  is the rating  for  50hz AC   ?

[radix]

Capacitors usually have a DC voltage rating, unless there is AC or ~ marked on them. That voltage equals to the peak voltage of an AC votage source. For a sine wave (frequency doesn't matter here) the peak voltage is U*sqrt(2). So basically your new cap is rated too low. Capacitors for connection to the mains (230V) usually have a rating of 250V~ or 400V.

You could also check the centrifugal starter switch. But it isn't necessary that there is one. Start capacitors usually have more capacitance.

[BradC]

Just measure the start and run windings to make sure they are both OK. Traditionally the start winding has a higher resistance than the main winding. If it's open then it is either burned out or there actually is a centrifugal start switch and it's jammed open. Caps don't last forever, although the old ones made domestically sure as hell last longer than the new asian sourced replacements.

[mrkev]

If you put two caps in series then unless you add in bleed resistors it's likely that most of the voltage gets dumped across one of them.

Sorry, but that's BS... This applies for DC voltages because Xc (reactance) is practically infinite. So the current going trough is determined by leakage. And since leakage can vary in order of decades (and changes over time), there can  (most likely) be x times more voltage drop on first cap then on the other.
But in AC, reactance dominates and current is determined by V*2*pi*f*C...

And for the OP, photos would be nice. First thing, use proper starter cap.  They are made for higher currents and for safety... I don't know what type of caps are you using now, but unless they are at least X2 safety rated, I wouldn't go near mains with them (and even if they are, it's not great idea)... Then, measure voltage over both coils (preferably with alligator clips so you are not near it when it's switched on). As many people suggested, the problem could be that you don't even get power to the second coil.

[Circlotron]

If it's a grinder then it's quite likely it doesn't have a centrifugal start switch because it starts with zero load. Stuff like (single phase) compressors that need big starting torque have such a switch to put a second cap in circuit momentarily. Then you have so-called split phase motors like often used in washing machines that use the switch to connect the starting winding but don't use a cap.

Your motor is almost certainly like this -> http://en.wikipedia.org/wiki/AC_motor#Permanent-split_capacitor_motor

[mushroom]

JacquesBBB, as I can see, you're french.

Do you know this website, created by a retired professionnal ?

https://sites.google.com/site/rebobinagebobinage/

It's by far the best information on the french web about electric motors. I've seen this guy fixing many problems on machine tools and other stuff, and even successfully teaching how to rewind motors.

There are two capacitor types: start and run. They are very different. Capacitors are not the same, depending of the motor : universal or asynchronous. Centrifugal means asynch, with one start and one run (as far as I remember...). Use the wrong type, and it can be destroyed (explosion). But it's funny to see a capacitor explosion (one of my games when I was a kid, when anal retentives did'nt say "don't play with 220V")

[KJDS]

If you put two caps in series then unless you add in bleed resistors it's likely that most of the voltage gets dumped across one of them.

Sorry, but that's BS... This applies for DC voltages because Xc (reactance)

You're right, I wasn't awake enough to sort out my AC and DC cases.

[JacquesBBB]

Thanks  to everyone for their inputs which  all of them have been very helpful.
I will make a tentative of summary of all the above inputs, which may be helpful for others.

First general matters
1) Capacitors in series.
Many things have  been said, and willing to clarify it, the best thing is to do the maths. I  have summarized  it in this little outline, showing at the same time  an example of circuitikz, an efficient drawing package for circuits in LateX.
I put both the source code and output.

So in summary,

for DC : V1=V R1/(R1+R2)

for AC : V1~ V C2/(C1+C2)


2) The caps voltage are given in DC, unless specified by "~".
From DC rating the AC  rating should be used considering the peak voltage, that is

AC rating = DC rating /sqrt(2)


2) Good  web sites for motors

http://www.repairfaq.org/REPAIR/F_appfaqh.html#APPFAQH_003

and in french

https://sites.google.com/site/rebobinagebobinage/
[\quote]

[JacquesBBB]

Now specific matters for my motor.

First some pictures



So you see, there does not seem to be a  centrifugal switch. The configuration is probably  a permanent Cap.


As I said,  the  original cap is a 4uF cap. Many have said that the value is small, but this is what it is.

[JacquesBBB]

Finally some measures, and  the schematics, as I understand it,
but I have not open yet the core of the  motor.




Again with the circuitikz file.

[JacquesBBB]

In addition, I have measured the caps after my test, and it is  4.545 uF,  with 0.14ohm ESR, so well within specs (4.7 uF), and not destroyed by the experiment.
As you see these are not recent caps.

I know this is not the proper cap, but  unless you provide evidence of the contrary, there should be no  difference for a short test to be sure that I can start this motor in  a normal way.  I will ultimately order  an adapted cap for  normal use of the motor.

Unfortunately, although I did not open the core of the motor, I see now that there is some  friction with it.  Maybe while blowing air to clean everything, I  blow some dust in the inside of the motor, and I will have now to completely tear it apart.

[mrkev]

OK, those caps look like they can handle this for some time. I was afraid you are using something less robust.

Anyway, resistance of Laux is suspiciously low; it's usually more than third of resistance of main coil, often even bigger than the resistance of main coil.
I would say, that there is a chance, that those blown caps took away the aux coil and it's now shorted (polyurethane insulation melted and there is contact between turns).

[JacquesBBB]

Thanks  for your message

I would say, that there is a chance, that those blown caps took away the aux coil and it's now shorted (polyurethane insulation melted and there is contact between turns).

I doubt this because the old capacitor was not shorted.  On the opposite, its resistance was a few Mohms.

I have dismounted the motor.  It is an induction AC motor with no other capacitors nor centrifugal switch.

I have cleaned everything, remove most of the rust, cleaned the bearings, which are just bronze olives, oiled them.
Remounted everything.
 
Again it starts when I hand start it, but not with the 4.7 uF cap, not with a similar 10 uF cap.

I  tested it out from its shell, and I found that it was heating a lot.

[wraper]

OK, those caps look like they can handle this for some time. I was afraid you are using something less robust.

Guess this some time will not be long at all. Even 400V DC capacitors do fail very often when exposed to 230V AC. Here is only 160V DC x2, however, because of the non perfect voltage distribution which happens because of the capacitance tolerance, it is even worse than that. Likely will last for the test time though.

[wraper]

Capacitors for connection to the mains (230V) usually have a rating of 250V~ or 400V.

Nope, that ~250V is not the same as the rating for usual capacitors. This is very tough load condition. Just some example from the datasheet of X2 ~250V rated capacitor: 800 VDC at 85 °C, 630 VDC at 110 °C.

[mrkev]

Thanks  for your message

I would say, that there is a chance, that those blown caps took away the aux coil and it's now shorted (polyurethane insulation melted and there is contact between turns).

I doubt this because the old capacitor was not shorted.  On the opposite, its resistance was a few Mohms.

That doesn't really mean anything... It could short out for a few seconds and than  disconnect due to vaporization of the metal inside that cap. Or the wiring is just old and mechanically stressed and the cap died because of it being shorted...

Again it starts when I hand start it, but not with the 4.7 uF cap, not with a similar 10 uF cap.
I  tested it out from its shell, and I found that it was heating a lot.

That sadly supports my previous theory, that the aux wiring is gone and you will have to rewind... Btw, does it heat up without that aux connected - running only at main coil?

[JacquesBBB]

That sadly supports my previous theory, that the aux wiring is gone and you will have to rewind... Btw, does it heat up without that aux connected - running only at main coil?

I think you are right !  Thanks !

I tried again without the capacitor, and this time the motor did not heat.

How difficult is it to rewind a motor ?

It there now  another way to make a motor start ?  For example with some electronic device in order to shift
the phases as in a DC brushless motor ?

[BradC]

How difficult is it to rewind a motor ?

It's not easy, but it's not particularly difficult, just time consuming and fussy.

It there now  another way to make a motor start ?  For example with some electronic device in order to shift
the phases as in a DC brushless motor ?

No, a DC brushless motor works because it has 3 identical windings 120 degrees out of phase. A PSC motor has two windings with the start winding usually 90 degrees out of phase. Without that winding you have exactly one winding and no way to generate an apparent phase shift.

Since the motor is toast, you could either take it to a motor re-winder and have it done, or get dirty and strip the windings to figure out how to wind new ones.
If you are lucky you should be able to just get the start winding out and re-construct that.

The other option is to get used to giving it a flick to start and just using it until it dies.

[mrkev]

How difficult is it to rewind a motor ?

Just look for motor rewinding on youtube, there is plenty of stuff. It's not that hard, but a really fiddly job requiring a lot of patience (especially getting those coils in the stator). And I don't know if you'll be able to impregnate it. Anyway, I would recommend to rewire whole thing, after all, it's old motor...
There could be some local company that does this, question is for how much, but I would rather use them...

It there now  another way to make a motor start ?  For example with some electronic device in order to shift
the phases as in a DC brushless motor ?

Not really. Main wiring of 1F motor doesn't make rotational magnetic field (f.e. it's going only up and down), the aux coil creates 90deg shifted field (sideways), without operational aux coil, you can't produce that effect.
On the other hand, this kind of operation isn't really damaging for the motor as long as it's spinning (you'll get slightly less power), in contrast with 3F motors, which produce magic smoke if you let them run at 2 phases...
You can even spin it by hand and then press the start switch (if you are not using it daily and don't wanna spend time/money, that's IMHO the best solution)...

[bills]

If the values on the schematic you posted on page 2 are correct you need to try real 4.7 uf  +/ - 20 %   run cap@ 370v check the windings to ground before you waist any more time.

[SeanB]

Start winding is cooked, you can see the discolouration in the photo under the varnish. Rewinding will cost around $100 ( price here for a small motor, the rewinder is near me in a small shop in a back alley, near the stadia) so if the motor is worth it to you, for sentimental or other reasons, take it to a local rewinder ( just the stator only) and ask for a quote.

On a positive note the rewind will come back with class H insulation, as opposed to the old stuff which dies at 130C. It will now survive another few decades.

[Circlotron]

It there now  another way to make a motor start ?  For example with some electronic device in order to shift
the phases as in a DC brushless motor ?

You really need both windings to function for this motor to run properly. In this particular motor the second winding that is powered through the capacitor is not just a starting winding, it is the second phase because the machine is actually a 2-phase motor. if you were to run it electronically you need to give it two sinewaves separated by 90 degrees, both windings always powered.

[Circlotron]

You can even spin it by hand and then press the start switch (if you are not using it daily and don't wanna spend time/money, that's IMHO the best solution)...

With the capacitor in circuit the current sent to the bad winding was limited. If you now use a start switch into what is now a near short circuit I think you might completely blow up the rest of the windings. 

Edit -> I thought you meant a "start" switch wired where the capacitor was.
Now I realise you mean the normal on-off switch.
But putting a switch in place of a cap and feeding a shorted winding would be a disaster.

[mrkev]

You can even spin it by hand and then press the start switch (if you are not using it daily and don't wanna spend time/money, that's IMHO the best solution)...

With the capacitor in circuit the current sent to the bad winding was limited. If you now use a start switch into what is now a near short circuit I think you might completely blow up the rest of the windings. 

Edit -> I thought you meant a "start" switch wired where the capacitor was.
Now I realise you mean the normal on-off switch.
But putting a switch in place of a cap and feeding a shorted winding would be a disaster.

Duh, I meant that it would run only on main coil. And OP wrote that he would start the motor by giving it a notch, which is stressful for wiring, if you wanna start motor like this with aux coil disconnected, it's better to spin it first, not after you switch it on...

[JacquesBBB]

Thanks all for your concern.

Once I realized, as mrkev said, that the aux was shorted, I disconnected the cap,  as indeed, this was much worse than with the old cap that had a high resistor value.

Unfortunately,  in my tests  the motor heated a lot,  and maybe now it is worse than in the original state.
I  could start it manually, but it would stop, and now does not want to start again.

So now I understand that the only way would be to rewind it,  but  it  is too much work. I will just leave it like that,
and provide to my friend my own grinder  which is a cheap chinese one on which he could put his stone.
And if I manage some day to find a electric motor of the same kind, I  will replace it,
as    I find the  casing very nice.

So it is an example of total failure in searching to fix something,  but at least I learned something.

Thanks to all for your contributions !